Exercise Addiction During the COVID-19 Pandemic: an … · 2020. 12. 3. · ORIGINAL ARTICLE...

ORIGINAL ARTICLE

Exercise Addiction During the COVID-19 Pandemic:an International Study Confirming the Needfor Considering Passion and Perfectionism

Ricardo de la Vega1 & Lucia Jiménez Almendros1 & Roberto Ruíz Barquín2 &Szilvia Boros3 & Zsolt Demetrovics4 & Attila Szabo3,4

Accepted: 11 November 2020/# The Author(s) 2020

AbstractVarious levels of lockdown due to COVID-19 limit people’s habitual physical activity.Individuals addicted to exercise, health-oriented, and team-exercisers could be the mostaffected. We examined the COVID-19-related changes in exercise volume in 1079exercisers from eight Spanish-speaking nations based on exercise addiction categories,primary reasons for exercise, and forms of exercise. The COVID-19-related decrease inexercise volume was 49.24% in the sample. The proportion of the risk of exerciseaddiction was 15.2%. Most (81.7%) of the participants exercised for a health-relatedreason. These exercisers reported lesser decrease in their exercise volume than thoseexercising for social reasons. The risk of exercise addiction was inversely related tochanges in exercise volume, but after controlling for passion and perfectionism therelationship vanished. The reported effect of COVID-19 on training did not differbetween the exercise addiction groups. The findings also confirm that exercise addictionresearch should control for passion and perfectionism.

Keywords Exercise dependence . Individual sport . Reason for exercise . Team sport . Training

A physically active lifestyle has physical (Lee et al. 2011) and mental health benefits (Clow andEdmunds 2013). Most often people engage in physical activity for a health reason, such as coping

International Journal of Mental Health and Addictionhttps://doi.org/10.1007/s11469-020-00433-7

* Attila [email protected]

1 Department of Physical Education, Sport and Human Movement, Autonomous University of Madrid,Madrid, Spain

2 Department of Developmental and Educational Psychology, Autonomous University of Madrid,Madrid, Spain

3 Institute of Health Promotion and Sport Sciences, ELTE Eötvös Loránd University, Bogdánfy u. 10/B, Budapest 1117, Hungary

4 Institute of Psychology, ELTE Eötvös Loránd University, Budapest, Hungary

http://crossmark.crossref.org/dialog/?doi=10.1007/s11469-020-00433-7&domain=pdfhttps://orcid.org/0000-0003-2788-4304mailto:[email protected]

with stress (Berczik et al. 2012; Szabo et al. 2019). In certain cases, the habitual physical activitycould become compulsive (Stevens et al. 2013), which stems from tolerance (a natural trainingeffect) and urges the person to progressively increase her/his exercise to achieve the same benefits asbefore. The craving for more and more exercise may result in loss of control, and from that point onthe behavior is no longer “healthy” (Szabo 2010).When a person loses control over her/his physicalactivity a dysfunctional behavior, known as “exercise addiction,” may be observed (Szabo 2010;Szabo et al. 2015). While this morbidity is relatively rare (i.e., ≈ 3.7%) as based on a recent meta-analysis (Trott et al. 2020), there are about 1000 published scholastic articles in the field (Szabo andKovacsik 2019). The keen interest in the topic was only recently justified by Juwono and Szabo(2020) by revealing that there might be significantly more than the currently estimated cases ofexercise addiction.

One overt sign of exercise addiction is the very high or exaggerated amount of training.However, MacLaren and Best (2007) noted that while exercise addiction is connected to highexercise volume, the latter is not evidence for addiction. Indeed, passion for exercise charac-terized by the love for the physical activity that one finds appealing, salient, and keenly investstime and energy into it (Vallerand et al. 2006) may be reflected in high volumes of exercise.Vallerand et al. (2003) have suggested that passion can be harmonious (HP) and obsessive(OP). The former is manifested when the beloved activity is autonomously internalized and theindividual engages in it with flexibility (Vallerand et al. 2003; Vallerand et al. 2006). On thecontrary, OP is manifested when the person internalizes the adopted activity in a controlledway and the activity is rigidly controlled (Vallerand et al. 2003). Further, the obsessivelypassionate person attaches importance to the activity’s corollaries, such as self-esteem orescape from stress, which renders it difficult to control or to terminate the passionate activity(Vallerand 2010). There is commonality between exercise addiction and passion. Indeed, OPexplains a large amount of variance in exercise addiction (Kovacsik et al. 2019, 2020). Theshared variance is more than twice as large in individual than in team exercisers (Kovacsiket al. 2020). Passion mediates the association between exercise addiction and exercise volume.In fact, it may wipe out the connection between the two (Szabo and Kovacsik 2019).Therefore, it was suggested that in studying exercise addiction, researchers must control forthe effects of passion (Szabo and Kovacsik 2019). Indeed, these authors questioned thevalidity of the results of all earlier studies on exercise addiction that did not control for passion.

Apart from the strong link between exercise addiction and passion, research shows thatperfectionism is associated with both (Bircher et al. 2017; Curran et al. 2014). Schipfer and Stoll(2015) in their Exercise-Addiction/Exercise-Commitment-Model (EACOM) suggested that passionand perfectionismmightmoderate exercise addiction. They produced preliminary evidence showingthat certain aspects of both passion and perfectionism do indeed moderate exercise addiction. Thus,research on exercise addiction should not only control for passion but also for perfectionism.

TheCOVID-19 affects thewholeworld. Thosewho exercise in large volumes, including those atrisk of exercise addiction and group/team exercisers, are forced to cut down on their exercise.However, individuals addicted to exercise will presumably find alternative means of training,including home or individual outdoor exercises that are permitted during the lockdown. Similarly,individuals who exercise for a health, or a “therapeutic” reason,may exert extra effort in trying not tocut down their volume of exercise, or otherwise theymay feel loss of control over the health ailmentfor which they exercise (Szabo et al. 2019). Considering the interactional model for exerciseaddiction (Egorov and Szabo 2013), individuals who get addicted to exercise have a therapeutic(health) primary motive related to coping with life difficulties. Therefore, people at high risk ofexercise addiction, exercising for a health reason,may not be able to cut down their exercise volume.

International Journal of Mental Health and Addiction

Further, individuals addicted to exercise in team or a group exercises might be forced to decreasetheir exercise volume during the COVID-19 pandemic, because the unavailability of open facilitiesand absence of other group members. Consequently, it is of interest to investigate how those at riskof exercise addiction cope with the COVID-19 situation in terms of exercise volume and perceivedeffect of the COVID-19 pandemic on their exercise training regimen. It is also of interest to furtherinvestigate this question in individuals at risk of exercise addiction who exercise admittedly fortherapeutic reason(s) and those who train in group or team settings.

The purpose of the study was to investigate changes in exercise patterns in the context of the riskof exercise addiction during early stages of the COVID-19 pandemic in eight Spanish-speakingnations, while also considering the roles of passion and perfectionism. Specifically, we wereinterested in examining the following: (1) the prevalence of the risk of exercise addiction in thesample; (2) the distribution of the primary exercise motive (i.e., (a)) health-, (b) skill-, and (c) social)in three exercise addiction risk groups (i.e., (a)) those at risk of exercise addiction [strong symptoms],(b) symptomatic [mild symptoms], and (c) asymptomatic [no symptoms]); (3) the change in exercisevolume, in response to the COVID-19 pandemic in the three exercise addiction risk groups byconsidering their primary reason for exercise; (4) the relationship between the change in exercisevolume during the COVID-19 pandemic and exercise addiction risk scores; (5) the change inexercise volume in response to the COVID-19 pandemic in three exercise addiction groups in thecontext of team or individual exercises; and (6) the perceived effect of the COVID-19 situation onthe habitual exercise training regimen in the three exercise addiction risk groups.

Apart from the first question that was exploratory, we posed five hypotheses: (1) the primarymotives for exercisewill differ between the three exercise addiction risk groups; (2) those at high risk ofexercise addiction having a primary therapeutic motive for exercise will exhibit less change in exercisevolume due to the COVID-19 situation than those who are at lower risk and exercise for other thanhealth reasons; (3) a negative relationship will emerge between the exercise addiction risk scores andthe change in exercise volume during the COVID-19 pandemic; (4) team exercisers in three exerciseaddiction risk categories will demonstrate greater change (decrease) in exercise volume than individualexercisers during the COVID-19 pandemic (due to lack of teammates and closed facilities); and (5)those at risk of exercise addiction will report greater negative impact of the COVID-19 pandemic intheir training schedules than symptomatic or asymptomatic exercisers. Due to the very novelty of theCOVID-19 literature, little or no background literature is available on the connection between thepandemic and the dependent measures. Therefore, the current hypotheses were generated primarily byrelying on “common sense” or rational expectations. However, to investigate the claim of Szabo andKovacsik et al. (2019) that the nearly 1000 published papers in the field of exercise addiction, whichdid not control for passion,mayhave reported unreliable results,we double tested our hypotheses, oncewithout including passion in the model and one more time by controlling for passion. Expandingfurther the work of Szabo and Kovacsik et al. (2019) and accounting for EACOM (Schipfer and Stoll2015), we also added perfectionism to the second model (conditionally, based upon a significantrelationship between exercise addiction risk scores and perfectionism).

Materials and Methods

Participants

Participants were recruited via social media from eight Spanish speaking countries: Argentina (n=102), Chile (n= 95), Costa Rica (n = 93), Ecuador (n= 99), Honduras (n = 77), Mexico (n= 96),


Spain (n= 422), and Uruguay (n= 95). The total number of eligible participants was 1079 (52%men). Their mean age was 32.88 (± SD=11.73) years, ranging from 18 to 75 years. More than half(53.1%) of the sample was involved in team/group exercises while the rest exercised on anindividual basis. The majority (81.7%) exercised primarily for a health-related reason, 11.7% fora skill-related reason, and 6.6% for a social reason. While most participants were involved in oneform of exercise, 24% of them reported practicing more than one exercise.

Materials

A battery of questionnaires, comprised of demographic and exercise habits questions as well asthree psychometric instruments, was used in the current study. The demographic questionsonly asked for age and gender. Questions about exercise habits gauged the main reason forexercise behavior (i.e., health, skill, or social relation related), the weekly amount of exercise(hours per week) before and during the COVID-19 pandemic, and form of exercise (individualor team). A single item 11-point Likert scale was used for assessing the perceived impact ofCOVID-19 on respondents’ training regimen. Additionally, three psychometric instrumentswere employed for measuring the risk of exercise addiction, passion, and perfectionism.

The Exercise Addiction Inventory (EAI; Terry et al. 2004) was used to assess the risk ofexercise addiction (EA). An example statement is “Exercise is the most important thing in mylife.” The EAI has six items which are rated on a 5-point Likert scale, ranging from “stronglydisagree” to “strongly agree.”Higher scores indicate higher risk of EA. The EAI was presentedwith good psychometric properties (Griffiths et al. 2015; Terry et al. 2004). In the currentresearch, we used the Spanish version of the scale (Sicilia et al. 2013). The internal reliabilityof the Spanish EAI is (Cronbach’s α) 0.70 (Griffiths et al. 2015; Sicilia et al. 2013). In thecurrent inquiry, the internal reliability of the EAI was 0.65, which fits the lower level of the αrange (α = 0.61 to α = 0.80) reported for the EAI in a cross-cultural investigation (Griffithset al. 2015).

Passion was assessed with the revised Passion Scale (PS; Marsh et al. 2013) whichmeasures harmonious passion (HP), obsessive passion (OP), on two 7-item subscales andpassion criteria (PC) on a 5-item subscale. The latter is usually used as control for the validityof OP and HP. Its five items reflect the conceptualization of passion (Vallerand et al. 2003).For example, they mirror how much a person likes or loves the activity, appreciates it,dedicates time and energy to it, views part of her- or himself, and considers the activitypassionate. While PC is not used as a dependent measure, it should be positively related toboth OP (i.e., in the present study r = 0.57, p < 0.001) and HP (r = 0.67, p < 0.001) (Marshet al. 2013; Vallerand et al. 2003). All its items are rated on a 7-point Likert scale, ranging from“not agree at all” to “very strongly agree”. A sample item for OP is “If I could, I would only domy activity”, for HP: “This activity is in harmony with the other activities in my life,” and forPC: “This activity is part of who I am.” Higher scores reflect higher passion on each subscale.Here, we used the Spanish version of the scale (Chamarro et al. 2015). The internal reliability(α) of the PS in the current work was 0.89, which is a higher value than that (0.81) reported byChamarro et al. (2015). The internal reliabilities of the three subscales were as follows: OP =0.83, HP = 0.82, and PC = 0.83. Given that it was repeatedly demonstrated that passion sharesa large proportion of the variance with EAI scores (Kovacsik et al. 2019, 2020), its assessmentwas used for controlling the moderating effects OP and HP on the risk of exercise addiction.

We measured perfectionism with Frost Multidimensional Perfectionism Scale (FMPS;Frost and Marten 1990). This scale can be used as both unidimensional and multidimensional


instrument. In the former case, 29 of 35 items are summed to provide an overall measure ofperfectionism. The scale’s six subscales1, of which five evaluate dimensions of perfectionism,such as personal standards or doubt about one’s action, could be used for measuring thevarious components of perfectionism separately. However, in this work, we were interested inthe moderating role of the unidimensional perfectionism on the risk of exercise addiction. Asample item is “I have extremely high goals.” All items are rated on a 5-point Likert scale. Sixitems, representing “organization” are not considered. After reversely rating some of the items,the individual ratings are summed up for 29 items to obtain a total score for perfectionism. Inthe present study, we employed the Spanish version of the FMPS (Gelabert et al. 2011). Itsinternal reliability in the current sample was 0.93.

Procedure

After obtaining ethical clearance from the Autonomous University of Madrid, demographicquestions and the three psychometric instruments were published on Google Forms surveyplatform. Volunteers were eligible to take part in the study if they were aged 18 years or overand exercised for at least 1 hour per week. The gathered data were checked and the answers ofthose who did not meet the criteria for participation were deleted from the dataset (n = 37).Prior to being able to see and answer the questionnaires, participants gave consent toparticipation by marking with a mouse- click an “agree” button at the end of the consentform. This form contained information about the study as well as criteria for participation. Datawere collected between the 6th and 20th of April 2020 and were saved in an Excel file.Subsequently, this file was imported in the SPSS statistical software (version 26) which wasused for data verification and analyses.

Data Reduction and Analyses

The grouping of the risk of exercise addiction was performed after rating the EAI. Accord-ingly, a score of 24 or above reflects risk of exercise addiction, a score between 13 and 23indicates moderate risk (symptomatic), and a score between 6 and 12 is indicative of low risk(asymptomatic) (Terry et al. 2004). Change in exercise volume was calculated by subtractingthe current total weekly hours of exercise from that reported for the period before the COVID-19 pandemic. Descriptive statistics were used in determining the overall prevalence of the riskof exercise addiction in the current sample. Chi square (χ2) was used to examine the frequencyof health, skill, and socially motivated exercise among the three exercise addiction risk groups.A bivariate and a partial correlation (controlling for passion [OP, HP] and perfectionism) werecalculated to examine the relationship between exercise addiction risk scores and changes inexercise volume. Analysis of covariances (ANCOVA) was used in examining the change inexercise volume, as a result of the COVID-19 pandemic, in individuals at risk of exerciseaddiction having different primary motives for exercising as well as in those exercising in teamand individual exercises. Further, an ANCOVA was also employed in examining the per-ceived impact of the COVID-19 pandemic on the training regimen of the three exerciseaddiction risk groups. In addition to gender and age, in all these analyses, we also controlledfor passion as suggested by Szabo and Kovacsik et al. (2019), as well as perfectionism.

1 A version with four subscales also exists, but the psychometric properties of the six-subscale Spanish versionare superior to that of four-subscale version.


Results

The overall decrease in habitual exercise was nearly 50% (49.24%) in the whole sample.Participants exercised an average of 9.22 (± SD = 5.65) hours/week before and 4.54 (± SD =3.76) hours/week during the COVID-19 pandemic (t[1078] = 29.84, p < 0.001, effect sizeCohen’s d = 0.91). The proportion of those at risk of exercise addiction was 15.2%, thoseshowing some symptoms amounted to 79.6%, and 5.2% were asymptomatic. The χ2 testyielded statistically no significant difference between the eight nations. However, a statisticallysignificant distribution emerged in the observed frequency of the three primary exercisemotives (health, skill, and social) in the three exercise addiction risk groups (χ2[4] = 9.88,p = 0.043). As illustrated in Table 1, this result may be attributed to the different ratio of skill,but mainly social, reason in the asymptomatic group compared to the other two groups.Among all exercise addiction risk groups, health was the most frequently reported motivefor exercise (81.7%).

To test the hypothesis that those at risk of exercise addiction, having a health-relatedprimary motive for their exercise, exhibit the least change in exercise volume due to theCOVID-19 pandemic, we calculated two ANCOVAs. The first only controlled for age andgender while the second also controlled for passion (both OP and HP). Further, because astatistically significant correlation emerged between EAI scores and perfectionism (r = 0.26,p < 0.001), we also controlled for perfectionism in the second model. The results of the firstmodel yielded a main effect for exercise addiction risk groups (F[21066] = 4.32, p = 0.013) andfor primary exercise reason (F[21066] = 5.74, p = 0.003), but no interaction between the two. Atthis early point, we disregarded the between-group differences, because we wanted to obtainthe results of the second ANCOVA which also included passion and perfectionism in themodel. This test was no longer statistically significant for the exercise addiction risk groups,(p > 0.05), but it was still significant for the primary reason for exercise (F[21061] = 4.55, p =0.011; see Fig. 1). The Bonferroni corrected post-hoc tests revealed that respondents exercisingfor health reasons decreased their exercise volume less than those exercising primarily for asocial reason (p = 0.018), but neither of these groups differed from those who exercisedprimarily for a skill-related reason.

To test the relationship between the risk of exercise addiction and the change in weeklyexercise volume, first we calculated a bivariate correlation, which was weak, but statisticallysignificant (r = −0.127, p < 0.001). However, when a partial correlation controlling for passionand perfectionism was applied, as suggested in the literature (Szabo and Kovacsik 2019), therelationship was no longer significant (r = −0.024, p > 0.05). The zero-order correlationsshowed that exercise addiction was positively and statistically significantly related to allcontrol variables (OP: r = 0.57, r2 = 0.32, p < 0.001; HP: r = 0.42, r2 = 0.18, p < 0.001; PC:r = 0.49, r2 = 0.24, p < 0.001; and perfectionism: r = 0.26, r2 = 0.07, p < 0.001). The result of

Table 1 Frequency of exercise motives in three exercise addiction risk groups

EAI Classification Health Skill Social

Asymptomatic (n = 56) 75% 8.9% 16.1%Symptomatic (n = 859) 82.7% 11.5% 5.2%At risk (n = 164) 79.3% 13.4% 7.3%Total (n = 1079) 81.7% 11.7% 6.6%


the partial correlation justified the consideration of the second model in ANCOVAs whichcontrolled for passion and perfectionism in the context of the risk of exercise addiction.

The ANCOVA, testing the hypothesis that team/group exercisers at various risk levels ofexercise addiction show greater decrease in exercise volume due to the COVID-19 pandemicthan individual exercisers, yielded a statistically significant interaction when controlling forage and gender (F[21069] = 3.80, p = 0.023), as well as when we added passion and perfection-ism to the model (F[21064] = 3.44, p = 0.033). This result was due to a significantly largerdecrease in asymptomatic team exercisers than asymptomatic individual exercisers (seeFig. 2).

The ANCOVA, testing the hypothesis that those at risk of exercise addiction perceive agreater negative impact of the COVID-19 situation on their usual training regimen than the twoother lower-risk groups, was statistically not significant (see Fig. 3).

Discussion

The main contributions of the present international study are (1) the COVID-19 pandemicforces habitual exercisers to decrease their usual exercise volume to about half of the usualamount; (2) the majority of exercisers have a health-related primary reason for exercising; (3)if passion and perfectionism are not included in the model, there is a significant difference inthe primary reason for exercise between the three exercise addiction risk groups, but when themodel controls for these variables, the difference does not emerge anymore; (4) individualsexercising primarily for a health-related reason decrease less the volume of their exerciseduring the COVID-19 pandemic than those exercising for social reason; (5) a weak correlationbetween the decreased exercises volume during the COVID-19 pandemic and the risk ofexercise addiction exists only if there is no control for passion; otherwise, the relationship isnot significant; (6) only team exercisers in the asymptomatic exercise addiction risk group

Fig. 1 Change (decrease) in the weekly hours of exercise in the three exercise addiction risk groups, exercisingprimarily for health, skill, or social reason, as a result of the COVID-19 pandemic. Note: The value 0 on the Y-axis is the reference point (i.e., usual amount of exercise before the COVID-19 situation). CI, confidence interval;EAI, exercise addiction inventory based classification


have decreased more their exercise volume than individual exercisers in the same risk group;(7) the three exercise addiction risk groups do not differ in their perceived impact of theCOVID-19 pandemic on their habitual exercises.

A decrease to about half in the usual exercise volume could be ascribed to the COVID-19pandemic as revealed by the results of this international study. At this time, there is no reportavailable in the literature to which these findings could be compared. However, the hereobserved overall change in the habitual exercise volume may be considered “reasonable”based on limited movement, distance keeping, and closure of many exercise facilities in manynations during the COVID-19 pandemic. The current results pave the road for research on thephysical and mental health consequences of this major cut-back in physical activity in analready sedentary world.

Fig. 2 Change (decrease) in the weekly hours of exercise in the three exercise addiction risk groups because ofthe COVID-19 pandemic. Note: The value 0 on the Y-axis is the reference point (i.e., usual amount of exercisebefore the COVID-19 situation). CI, confidence interval; EAI, exercise addiction inventory based classification

Fig. 3 Perceived impact of the COVID-19 pandemic on the physical exercise training in three exercise addictionrisk groups. Note: CI, confidence interval: EAI, exercise addiction inventory based classification


The decrease in exercise volume was initially statistically significant between the threeexercise addiction risk groups, but when passion and perfectionism were added to the model,the three groups did not differ anymore. These findings support the strong connection betweenthe concepts of risk of exercise addiction and passion (Kovacsik et al. 2019, 2020) as well asperfectionism (Bircher et al. 2017; Curran et al. 2014; Schipfer and Stoll 2015). Therefore, theresults justify the recent call for considering passion, and based on the current study perfec-tionism too, in all research in which the risk of exercise addiction is a dependent measure(Szabo and Kovacsik 2019).

The decrease in exercise volume due to the COVID-19 situation was lower in health-oriented exercisers compared to those exercising primarily for a social reason. This findingonly partially confirms our hypothesis, because people exercisingmainly for a health reason didnot differ from the skill-oriented exercisers. The least cut down in exercise volume in this groupmight be explained in light of theHealth BeliefModel (Rosenstock 1974). The gist of themodelis that when a person feels threatened by a morbidity, which is determined by susceptibility andseverity, and benefits need to be evaluated against the barriers, greater effort is invested inconquering the barriers when the perceived threat is greater. During the COVID-19 pandemic,exercise facilities and social exercising may be limited, but if there is a threat to the health, theperson could find alternative ways to exercise. Therefore, in accord with our hypothesis, thisgroup manifests the least decrease in exercise volume (i.e., overcomes the most the imposedbarriers). Clearer distinction between health-motivated and other exercise reasons may becomemore evident over a longer period that can be investigated in longitudinal research.

In accord with our hypothesis, team exercisers decreased more their exercise than individualexercisers. However, when examining the decrease between the two groups in the context of therisk of exercise addiction, only asymptomatic team exercisers differed significantly from asymp-tomatic individual exercisers, with the former decreasing more than twice as much their exercisevolumes due to COVID-19 pandemic than the latter. There was almost negligible difference (seeFig. 2) in the amount of decrease in exercise volumes in the other exercise addiction risk groups. Itis possible that “usual” team exercisers who show some symptoms of risk of exercise addiction, orare at risk, compensate with individual exercises for the lost team exercise during the COVID-19crisis. This explanation should be tested empirically in future studies. Regrettably, we did not askthe types of exercise performed before and during COVID-19.

Despite our hypothesis that those at risk of exercise addiction will perceive a greater negativeconsequence of the COVID-19 situation on their training than the others, the results did notsupport our conjecture. In fact, there was no difference in the perceived impact of the COVID-19 situation between the three exercise addiction groups (see Fig. 3). All groups reported aslightly negative perceived effect (the 5 is the median of the assessment scale). We interpretthese findings as an outcome of a subjective appraisal of control conveyed in the responses,which is basically similar to all regular exercisers, regardless of their risk of exercise addiction.

Our hypothesis that there will be a negative relationship between the risk of exerciseaddiction and change in exercise volume during COVID-19 was supported only until wedid not control for passion and perfectionism. After controlling for these variables, therelationship vanished, and all zero-order correlations were significant, supporting the largecommonality between these factors (Szabo and Kovacsik 2019). Therefore, the current results,both those based on ANCOVA and correlation, lend support for the strong need to control forpassion in exercise addiction research as suggested by Szabo and Kovacsik. However, ourfindings expand their results by showing that perfectionism is also instrumental in the risk ofexercise addiction, and exerting control for this variable is also recommended. While the


correlation between exercise addiction and perfectionism was statistically significant, thecoefficient of determination (r2) was lower than the values obtained for passion, indicatingthat perfectionism shares less common variance with exercise addiction than passion. Never-theless, the connection between exercise addiction and perfectionism seems to affect the resultsrelated to exercise addiction.

Strengths and Limitations

The strength of the current work is that it is a pioneering study during COVID-19 in thecontext of exercise addiction reason for exercise and form of exercise, and it is based on largenumber of responses from eight Spanish-speaking nations. Still, another strength of the studyis that it is the first, based on authors’ best knowledge, to examine the risk of exercise addictionin the context of primary reason for exercise. Last, but not least, the results of the studyconfirm in a large sample that the inclusion of passion and perfectionism in the evaluationmodels of the risk of exercise addiction is necessary, because without controlling for passiondifferent results, encompassing different (but false) explanations, emerge. The limitations ofthe present study are manifested in the volunteerism, which is a general problem in psycho-logical studies. The internal reliability of the EAI was 0.65, in the current study which is at thelower end of the acceptable range and, hence, this is another limitation of this work. Further,these results cannot be generalized to non-Spanish speaking samples and despite all eightnations were affected by COVID-19, not all of them were affected equally. Indeed, the level oflockdown differed between these nations (BBC.com, April 6, 2020) and due to the anonymousnature of the inquiry, we do not know from which exact local regions did the people respond,which is important because within the eight countries examined there were regions more orless affected by the pandemic. Therefore, our findings may be more pronounced in the heavilyvirus-affected areas whereas they may be less evident in regions less affected by COVID-19.

Conclusions

Exercise volume decreases about 50% in regular exercisers 2 months into the COVID-19crisis. The decrease and the primary exercise motive are unrelated to the risk of exerciseaddiction. Most exercisers report a health-related motive as the primary reason for theirexercise and this motive is associated with lesser reduction of exercise volume than a socialreason. During COVID-19 team exercisers decrease more their exercise volume than individ-ual exercisers, but the difference is evident in the symptomatic exercise addiction risk grouponly. The findings confirm the call for controlling passion in exercise addiction research. Theyexpand this call by showing that perfectionism could also moderate the results obtained onexercise addiction and, therefore, this variable should be also considered, along with passion,in future exercise addiction research.

Funding Open access funding provided by Eötvös Loránd University.

Compliance with Ethical Standards

Conflict of Interest The authors have no conflict of interest to declare.


Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, whichpermits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, andindicate if changes were made. The images or other third party material in this article are included in the article'sCreative Commons licence, unless indicated otherwise in a credit line to the material. If material is not includedin the article's Creative Commons licence and your intended use is not permitted by statutory regulation orexceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copyof this licence, visit http://creativecommons.org/licenses/by/4.0/.

References

BBC.COM (2020). Coronavirus: The world in lockdown in maps and charts. https://www.bbc.com/news/world-52103747

Berczik, K., Szabó, A., Griffiths, M. D., Kurimay, T., Kun, B., Urbán, R., & Demetrovics, Z. (2012). Exerciseaddiction: Symptoms, diagnosis, epidemiology, and etiology. Substance Use & Misuse, 47(4), 403–417.https://doi.org/10.3109/10826084.2011.639120.

Bircher, J., Griffiths, M. D., Kasos, K., Demetrovics, Z., & Szabo, A. (2017). Exercise addiction and personality:A two-decade systematic review of th empirical literature (1995–2016). Baltic Journal of Sport and HealthSciences, 3(106), 19. https://doi.org/10.33607/bjshs.v3i106.30.

Chamarro, A., Penelo, E., Fornieles, A., Oberst, U., Vallerand, R. J., & Fernández-Castro, J. (2015).Psychometric properties of the Spanish version of the passion scale. Psicothema, 27(4), 402–409. https://doi.org/10.7334/psicothema2015.80 http://www.psicothema.com/psicothema.asp?id=4281.

Clow, A., & Edmunds, S. (2013). Physical activity and mental health. Champaign, IL: Human Kinetics. https://www.bookdepository.com/Physical-Activity-Mental-Health-Angela-Clow/9781450434331

Curran, T., Hill, A. P., Jowett, G. E., & Mallinson-Howard, S. H. (2014). The relationship between multidi-mensional perfectionism and passion in junior athletes. International Journal of Sport Psychology, 45(4),369–384 https://ray.yorksj.ac.uk/id/eprint/860/1/Curran,%20Hill,%20Jowett%20&%20Mallinson%202014.pdf.

Egorov, A. Y., & Szabo, A. (2013). The exercise paradox: An interactional model for a clearer conceptualizationof exercise addiction. Journal of Behavioral Addictions, 2(4), 199–208. https://doi.org/10.1556/JBA.2.2013.4.2 http://real.mtak.hu/55393/.

Frost, R. O., & Marten, P. A. (1990). Perfectionism and evaluative threat. Cognitive Therapy and Research,14(6), 559–572. https://doi.org/10.1007/bf01173364.

Gelabert, E., García-Esteve, L., Martín-Santos, R., Gutiérrez, F., Torres, A., & Subirà, S. (2011). Psychometricproperties of the Spanish version of the Frost multidimensional perfectionism scale in women. Psicothema,23(1), 133–139 https://pubmed.ncbi.nlm.nih.gov/21266154/.

Griffiths, M. D., Urbán, R., Demetrovics, Z., Lichtenstein, M. B., de la Vega, R., Kun, B., Ruiz-Barquín, R.,Youngman, J., & Szabo, A. (2015). A cross-cultural re-evaluation of the exercise addiction inventory (EAI)in five countries. Sports Medicine-Open, 1(1), 1–7. https://doi.org/10.1186/s40798-014-0005-5.

Juwono, I. D., & Szabo, A. (2020). 100 cases of exercise addiction: More evidence for a widely researched butrarely identified dysfunction. International Journal of Mental Health and Addiction. https://doi.org/10.1007/s11469-020-00264-6.

Kovacsik, R., Griffiths, M. D., Pontes, H. M., Soós, I., de la Vega, R., Ruíz-Barquín, R., Demetrovics, Z., &Szabo, A. (2019). The role of passion in exercise addiction, exercise volume, and exercise intensity in long-term exercisers. International Journal of Mental Health and Addiction, 17(6), 1389–1400. https://doi.org/10.1007/s11469-018-9880-1.

Kovacsik, R., Soós, I., de la Vega, R., Ruíz-Barquín, R., & Szabo, A. (2020). Passion and exercise addiction:Healthier profiles in team than in individual sports. International Journal of Sport and Exercise Psychology,18(2), 176–186. https://doi.org/10.1080/1612197x.2018.1486873.

Lee, D., Sui, X., Ortega, F. B., Kim, Y. S., Church, T. S., Winett, R. A., et al. (2011). Comparison of leisure-timephysical activity on cardiorespiratory fitness as predictors of all-cause mortality in men and women. BritishJournal of Sports Medicine, 46(6), 504–510. https://doi.org/10.1136/bjsm.2009.066209.

MacLaren, V. V., & Best, L. A. (2007). Symptoms of exercise dependence and physical activity in students.Perceptual and Motor Skills, 105(3 suppl), 1257–1264. https://doi.org/10.2466/pms.105.7.1257-1264.

Marsh, H. W., Vallerand, R. J., Lafrenière, M.-A. K., Parker, P., Morin, A. J. S., Carbonneau, N., & Paquet, Y.(2013). Passion: Does one scale fit all? Construct validity of two-factor passion scale and psychometric


https://doi.org/https://www.bbc.com/news/world--52103747https://www.bbc.com/news/world--52103747https://doi.org/10.3109/10826084.2011.639120https://doi.org/10.33607/bjshs.v3i106.30https://doi.org/10.7334/psicothema2015.80https://doi.org/10.7334/psicothema2015.80http://www.psicothema.com/psicothema.asp?id==4281https://www.bookdepository.com/Physical--Activity--Mental--Health--Angela--Clow/9781450434331https://www.bookdepository.com/Physical--Activity--Mental--Health--Angela--Clow/9781450434331https://ray.yorksj.ac.uk/id/eprint/860/1/Curran,%20Hill,%20Jowett%20&%20Mallinson%202014.pdfhttps://ray.yorksj.ac.uk/id/eprint/860/1/Curran,%20Hill,%20Jowett%20&%20Mallinson%202014.pdfhttps://doi.org/10.1556/JBA.2.2013.4.2https://doi.org/10.1556/JBA.2.2013.4.2http://real.mtak.hu/55393/https://doi.org/10.1007/bf01173364https://pubmed.ncbi.nlm.nih.gov/21266154/https://doi.org/10.1186/s40798-014-0005-5https://doi.org/10.1007/s11469-020-00264-6https://doi.org/10.1007/s11469-020-00264-6https://doi.org/10.1007/s11469-018-9880-1https://doi.org/10.1007/s11469-018-9880-1https://doi.org/10.1080/1612197x.2018.1486873https://doi.org/10.1136/bjsm.2009.066209https://doi.org/10.2466/pms.105.7.1257-1264

invariance over different activities and languages. Psychological Assessment, 25(3), 796–809. https://doi.org/10.1037/a0032573.

Rosenstock, I. M. (1974). Historical origins of the health belief model. Health Education Monographs, 2(4),328–335. https://doi.org/10.1177/109019817400200403.

Schipfer M., Stoll O. (2015). OR-77: Exercise-addiction/exercise-commitment-model (EACOM). Abstractpresented at the 2nd international conference on behavioral addictions — March 16–18, 2015, Budapest,Hungary, Switzerland. Journal of Behavioral Addictions, 4(S1), 35–37. https://akjournals.com/view/journals/2006/4/Supplement-1/article-p1.xml

Sicilia, Á., Alías-García, A., Ferriz, R., & Moreno-Murcia, J. A. (2013). Spanish adaptation and validation of theexercise addiction inventory (EAI). Psicothema, 25(3), 377–383. https://doi.org/10.7334/psicothema2013.21.

Stevens, R. E., Loudon, D. L., Yow, D. A., Bowden, W. W., & Humphrey, J. H. (2013). Stress in collegeathletics: Causes, consequences, coping. London, UK: Routledge https://www.routledge.com/Stress-in-College-Athletics-Causes-Consequences-Coping/Stevens-Loudon-Yow-Bowden-Humphrey/p/book/9780789009357.

Szabo, A. (2010). Addiction to exercise: A symptom or a disorder? New York: Nova Science Publishers.Szabo, A., Boros, S., & Bősze, J. P. (2019). Are there differences in life-satisfaction, optimism, pessimism and

perceived stress between therapeutic and mastery exercisers? A preliminary investigation. Baltic Journal ofSport and Health Sciences, 3(114). https://doi.org/10.33607/bjshs.v3i114.807.

Szabo, A., Griffiths, M. D., Marcos, R. D. L. V., Mervó, B., & Demetrovics, Z. (2015). Focus: Addiction:Methodological and conceptual limitations in exercise addiction research. The Yale Journal of Biology andMedicine, 88(3), 303–308 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4553651/.

Szabo, A., & Kovacsik, R. (2019). When passion appears, exercise addiction disappears. Swiss Journal ofPsychology, 78(3–4), 137–142. https://doi.org/10.1024/1421-0185/a000228.

Terry, A., Szabo, A., & Griffiths, M. (2004). The exercise addiction inventory: A new brief screening tool.Addiction Research and Theory, 12(5), 489–499. https://doi.org/10.1080/16066350310001637363.

Trott, M., Jackson, S. E., Firth, J., Jacob, L., Grabovac, I., Mistry, A., et al. (2020). A comparative meta-analysisof the prevalence of exercise addiction in adults with and without indicated eating disorders. Eating andWeight Disorders - Studies on Anorexia, Bulimia and Obesity. (online first). https://doi.org/10.1007/s40519-019-00842-1.

Vallerand, R. J. (2010). On passion for life activities: The dualistic model of passion. In M. P. Zanna & J. M.Olson (Eds.), Advances in experimental social psychology (pp. 97–133). San Diego, CA: Academic Presshttps://www.sciencedirect.com/science/article/pii/S0065260110420031.

Vallerand, R. J., Blanchard, C., Mageau, G. A., Koestner, R., Ratelle, C., Leonard, M., Gagne, M., & Marsolais,J. (2003). Les passions de l’ame: On obsessive and harmonious passion. Journal of Personality and SocialPsychology, 85(4), 756–767. https://doi.org/10.1037/0022-3514.85.4.756.

Vallerand, R. J., Rousseau, F. L., Grouzet, F. M. E., Dumais, A., Grenier, S., & Blanchard, C. M. (2006). Passionin sport: A look at determinants and affective experiences. Journal of Sport and Exercise Psychology, 28(4),454–478. https://doi.org/10.1123/jsep.28.4.454.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps andinstitutional affiliations.


https://doi.org/10.1037/a0032573https://doi.org/10.1037/a0032573https://doi.org/10.1177/109019817400200403https://akjournals.com/view/journals/2006/4/Supplement--1/article--p1.xmlhttps://akjournals.com/view/journals/2006/4/Supplement--1/article--p1.xmlhttps://doi.org/10.7334/psicothema2013.21https://doi.org/10.7334/psicothema2013.21https://www.routledge.com/Stress--in--College--Athletics--Causes--Consequences--Coping/Stevens--Loudon--Yow--Bowden--Humphrey/p/book/9780789009357https://www.routledge.com/Stress--in--College--Athletics--Causes--Consequences--Coping/Stevens--Loudon--Yow--Bowden--Humphrey/p/book/9780789009357https://www.routledge.com/Stress--in--College--Athletics--Causes--Consequences--Coping/Stevens--Loudon--Yow--Bowden--Humphrey/p/book/9780789009357https://doi.org/10.33607/bjshs.v3i114.807https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4553651/https://doi.org/10.1024/1421-0185/a000228https://doi.org/10.1080/16066350310001637363https://doi.org/10.1007/s40519-019-00842-1https://doi.org/10.1007/s40519-019-00842-1https://www.sciencedirect.com/science/article/pii/S0065260110420031https://doi.org/10.1037/0022-3514.85.4.756https://doi.org/10.1123/jsep.28.4.454

Exercise...AbstractMaterials and MethodsParticipantsMaterialsProcedureData Reduction and Analyses

ResultsDiscussionStrengths and Limitations

ConclusionsReferences

Exercise Addiction During the COVID-19 Pandemic: an … · 2020. 12. 3. · ORIGINAL ARTICLE...

Documents

Transcript of Exercise Addiction During the COVID-19 Pandemic: an … · 2020. 12. 3. · ORIGINAL ARTICLE...